Applications of weak attraction theory in \(\mathrm {Out}(\mathbb {F}_N) \) (Q280113)

Let \(F_{n}\) be a free group of rank \(n\geq 3\) and \(\mathrm{Out}(F_{n})\) its group of outer automorphisms. In this paper, the author studies a category of outer automorphisms of \(F_{n}\). Before quoting the main theorem, let's give a definition. Definition. Let \(\varphi, \psi \in\mathrm{Out}(F_{n})\) be an exponentially growing outer automorphisms, which are not conjugate to powers of each other and which do not have a common periodic free factor. Also let \(\varphi\), \(\psi\) have some dual lamination pairs \(\Lambda _{\varphi }^{\pm}\) and \(\Lambda _{\psi }^{\pm}\), respectively, such that the following holds: -- \(\varphi ^{\epsilon}\) leaves \(\Lambda _{\varphi }^{\epsilon}\) invariant and \(\psi ^{\epsilon}\) leaves \(\Lambda _{\psi }^{\epsilon}\) invariant. -- \(\Lambda _{\psi }^{\pm}\) is attracted to \(\Lambda _{\varphi }^{\pm}\) under iterates of \(\varphi ^{\epsilon}\) and \(\Lambda _{\varphi }^{\pm}\) is attracted to \(\Lambda _{\psi}^{\pm}\) under iterates of \(\psi ^{\epsilon}\). -- \(\{ \Lambda _{\varphi } ^{\pm} \}\cup \{\Lambda _{\psi }^{\pm}\}\) fills. -- Both \(\Lambda _{\varphi }^{\pm}\) and \(\Lambda _{\psi }^{\pm}\) are non-geometric or every lamination pair of every element of \(\langle \psi , \varphi \rangle \) is geometric. -- The non-attracting subgroup systems \(\mathcal{A}_{\mathrm{na}}( \Lambda _{\varphi }^{\pm})\) and \(\mathcal{A}_{\mathrm{na}}( \Lambda _{\psi}^{\pm})\) are mutually malnormal. Any two pairs \((\varphi , \Lambda _{\varphi }^{\pm})\), \((\psi , \Lambda _{\psi }^{\pm})\) wich satisfy the assumptions above will be called pairwise independent. Theorem A. Let \(\varphi\), \(\psi \) be two exponentially growing elements of \(\mathrm{Out}(F_{n})\), such that there exists some dual lamination pairs \(\Lambda _{\varphi }^{\pm}\) and \(\Lambda _{\psi }^{\pm}\), which makes \(\varphi\), \(\psi \) pairwise independent. Then there exists an \(M\geq 0\) such that for all \(n, m\geq M\), the group \(\langle \psi ^{m} , \varphi ^{n}\rangle \) will be free of rank two and every element of this free group, not conjugate (in \(G\)) to some power of the generators, will be hyperbolic and fully-irreducible. Corollary. Let \(\varphi\), \(\psi \) are fully-irreducible elements of \(\mathrm{Out}(F_{n})\) which are not conjugate to powers of each other, then there exists an integer \(M\geq 0\) such that for every \( n, m\geq M\), \(G = \langle \psi ^{m} , \varphi ^{n} \rangle \) is a free group of rank two, all whose elements are fully-irreducible. Moreover, \(M\) can be chosen such that: {\parindent=0.6cm \begin{itemize}\item[1.] If both \(\varphi\) and \(\psi \) are hyperbolic, then every element of \(G\) is hyperbolic. \item[2.] If \(\psi \) is hyperbolic and \(\varphi \) is geometric, then every element of \(G\) not conjugate to a power of \(\varphi \) is hyperbolic. \item[3.] If both \(\varphi\) and \(\psi \) are geometric but not fix the same conjugacy class, then every element every element of \(G\) not conjugate to a power of \(\varphi\) or \(\psi\) is hyperbolic. \end{itemize}} The first statement of this corollary is a theorem proved in [\textit{I. Kapovich} and \textit{M. Lustig}, J. Topol. Anal. 2, No. 2, 173--201 (2010; Zbl 1211.20027)] but with a very different technique. Theorem B: Let \(\mathcal{S}\) be a connected, compact surface (not necessarily oriented) with one boundary component. Let \(f, g\in \mathcal{MCG(S)}\) be pseudo-Anosov homeomorphisms of the surface which are not conjugate to powers of each other. Then there exists some integer \(M\) such that the group \(G=\langle f ^{m} , g^{n} \rangle \) will be free for every \(m, n\geq M\), and every element of this group will be pseudo-Anosov. This theorem is similar to a theorem by \textit{B. Farb} and \textit{L. Mosher} [Geom. Topol. 6, 91--152 (2002; Zbl 1021.20034)] except that here the surfaces have one boundary component. The used arguments and techniques were developed in previous papers concerning the study of \(\mathrm{Out}(F_{n})\). Here, we refer only to [``Subgroup decomposition in \(\mathrm{Out}(F_{n})\). III: Weak attraction theory'', Preprint, \url{arXiv:1306.4712}]. There, \textit{M. Handel} and \textit{L. Mosher} proved a proposition, which is strengthened by the author (Proposition 4.4 in the paper).